In rotating, non-removable data disc storage devices, a plurality of recording heads are provided, generally of the order of one head per disc surface. Each recording head has an electromagnetic transducer that records data on to a magnetic disc in concentric, circular tracks. It is desirable to provide the disc storage device with high data storage capacity, and one way of providing high capacity is to increase the number of discs in the storage device, generally achieved by manufacturing the discs to a very reduced thickness. It is also desirable to record data at high densities through the means of high frequency data patterns, high track densities, or both. It is also desirable to transfer data from the disc storage device to the host computer at as high a transfer rate as possible. Clearly, one way of increasing the data transfer rate is to rotate the discs at a very high frequency. In order to prevent abrasive wear of the head, or disc surface, or both, the recording head is provided with a self-acting hydrodynamic air bearing. The recording head is also provided with gimbal/suspension structure, which allows the head to freely follow undulations of the disc surface, a condition exacerbated by reduced disc thickness, without interfering with the performance of the air bearing, while at the same time maintaining accurate radial alignment of the electromagnetic transducer to the data track.
Considering the exemplary head/gimbal assembly (HGA) shown in FIGS. 1 and 2 supported over a disc surface segment, it is desirable to maintain rigidity of the slider 11 in the yaw and in-plane directions of the disc 13, while being resilient in the vertical, and pitch and roll directions to enable tracking of the topography of the disc. This is accomplished by the head/gimbal assembly (HGA) 14 of the hard disc file which consists of three components; the slider 11, the gimbal 15 and the load beam 16.
As set forth above, the slider 11 flies upon an air cushion or self-acting hydrodynamic bearing in extremely close proximity to the disc data surface 12. Further, the slider supports the electromagnetic transducer 10 for recording and retrieving computer data from a spinning magnetic disc. Electrical signals are sent and received from the transducer via very small, twisted copper wires.
The gimbal 15, which is coupled to the slider 11, provides resiliency to the slider's pitch and roll directions allowing the slider to follow the topography of the disc, while being rigid in the yaw and in-plane directions for maintaining precise slider positioning relative the data surface.
Finally, the load beam 16, which couples the gimbal 15 to a support arm 17, is resilient in the vertical direction to, again, allow the slider to follow the topography of the disc, and is rigid in the in-plane direction for precise slider positioning relative the data surface. The load beam 16 also supplies a downward force to counterbalance the hydrodynamic lifting force developed by the slider's air bearing.
Together, the load beam and gimbal comprise an assembly generally known as a head suspension or a head flexure, or simply as a flexure 18. Typical of these patented flexures are disclosed in U.S. Pat. No. 4,167,765 to Watrous and U.S. Pat. No. 4,245,267 to Herman.
In flexures such as in the '765 patent, a load point 21 is formed in the slider bonding surface of the gimbal. The downward force caused by the load beam 16 is applied directly to the load point 21. In order to have the slider fly above the disc with the correct attitude, it is more important that the forces and moments created by the hydrodynamic air bearing be properly balanced into equilibrium. It has been found that applying the load beam force to the slider via a gimbal load point guarantees proper flying attitude, since the balance point for the slider's air bearing is well-defined.
Historically, the gimbal 15 and load beam 16 have been fabricated discretely by chemically etching 300 series stainless steel foil into the desired shape. Subsequently, these discrete pieces are coupled together by laser welding or the like. While this arrangement has proven adequate to fabricate the individual pieces, other problems have arisen as a result of the etching process.
Briefly, in the chemical etching process, the stainless steel foil is first coated with photoresist which is a light sensitive liquid polymer. Initially, the foil sheets are dipped into a liquid photoresist bath for a predetermined amount of time to achieve a desired film thickness. Once the photoresist is of the desired thickness, the next step performed is known as patterning or developing which entails registering a photo mask over each side of the foil, and then exposing the mask to a light source. This exposure affects the solubility of the photo-sensitive components of the resist to developer solutions.
After completing the development step, the foil sheet will then be composed of areas of exposed stainless steel and areas of photoresist, and will be ready for chemical etching. In this process, the foil is exposed to an etchant such as ferric chloride. Typically ferric chloride is sprayed onto the foil for a controlled amount of time, and then the foil is rinsed. Following rinsing the photoresist is stripped by use of a special chemical solvent.
The chemical etching process is well known for producing very detailed parts having very crisp and well defined features. During etching, some of the ferric chloride undercuts the photoresist pattern such that a sharp edge or corner, such as edge 19, is formed between the two adjacent or common edges which intersect at an angle between 60 and 90 degrees. While these sharp edges 19 of the etched part are desirable from the point-of-view of consistency and repeatability, these edges are problematic in several respects.
The primary problem with sharp edges 19 is that the wires 20 from the transducer 10 to the amplifier must, unfortunately and unavoidably, route near or around one or more of such sharp edges. Since the copper wire is softer than the stainless steel of the flexure, the transducer wires can be easily nicked or cut if they come in contact with an etched edged. This is a continual cause of manufacturing defects of HGAs, and an occasional cause of head failures in hard disc files out in the field.
Another problem with two-piece flexures is the high cost of manufacture. Welding the two discrete components requires precision alignment fixturing, and careful handling of the delicate components in and out of the fixturing, since the parts are easily damaged. Welding also requires a well-controlled laser system that is capable of producing strong welds with minimal distortion to the structure from residual stress.
The fabrication of one-piece flexure assemblies often eliminate the load point feature to simplify manufacture. Typical of these one-piece flexures is described in the '267 patent. Another example is a one-piece flexure known as HTI Type 16, or T-16, manufactured by Hutchinson Technology, Inc. Neither of these one-piece flexures includes a well-defined load point. As a result, the flying height of sliders attached to these flexures vary widely and undesirably during operation.
One unitary flexure having a load point is described in U.S. Ser. No. 07/975,352 filed Nov. 12, 1992 entitled "ONE-PIECE FLEXURE FOR SMALL MAGNETIC HEADS" by Hagen, and U.S. Serial No. 08/432,843 filed May 2, 1995 entitled "FLEXTURE WITH REDUCED UNLOAD HEIGHT FOR HARD DISC DRIVE HEADS" by Hagen. These flexures have what is known as an etched load point to provide the well-defined balance point. Local to the etched load point on both inventions is some degree of out-of-plane cold forming to elevate the load point relative to the rest of the flexure. The cold forming process is well known, consisting of a precision male and female stamping dies that receive flat components, and stress the components beyond the yield point at selected locations. Because the forming process has a high degree of variance, it is generally necessary to target the elevation of the load point after forming sufficiently high to guard against worse case conditions. But this is disadvantageous since the overall height of the flexure is also increased, which ultimately limits the reduction of the spacing between adjacent discs in a multiple disc drive.
Moreover, since the load points of these flexures described in the above incorporated applications have been made by etching, the inner perimeter of the load point plateau is, of course, a sharp edge. The sliders are generally made from aluminum-oxide titanium-carbide which is harder than stainless steel, and therefore resistant to damage from the flexure. Nevertheless, the sharp edges of the etched load point remain a concern.
A further problem with the current flexure manufacturing process relates to the control of static attitude, which is the orientation of the surface that the slider is bonded to relative to other principal flexure datums. It has been found that one of principal factors causing variation in flying height is the variation in static attitude. Moreover, recent studies have determined that the static attitude variance of the head/gimbal assembly (HGA) is primarily caused by the static attitude variance of the flexures.
Verifying the static attitude on HGAs can easily be done by reflecting a helium-neon laser beam against the slider's air bearing surface while the HGA is clamped to a reference datum. The HGA's air bearing surface is polished to a mirror finish which enables the laser to reflect with little dispersion of the beam.
Unfortunately, inspecting static attitude variance of the flexure is more difficult than with the HGA This is due to the fact that the finish of the flexure is much more coarse since the foils employed to fabricate the flexure have a rolled finish with grain lines. Thus, the inspection of flexure static attitude requires more complicated equipment, that often has poor through-put. Further, as the roughness of the surface increases, the measurement to determine the plane of a surface becomes more uncertain and less reproducible.